A Photochemical, Room-Temperature, and Aqueous Route to the

Many different pathways have been reported in the literature to account for the .... of a typical nanocrystal sample, matching a pure fcc Pd spectrum ...
3 downloads 0 Views 9MB Size
Download PDF
Article pubs.acs.org/cm

A Photochemical, Room-Temperature, and Aqueous Route to the Synthesis of Pd Nanocubes Enriched with Atomic Steps and Terraces on the Side Faces Madeline Vara,† Ping Lu,‡ Xuan Yang,‡ Chi-Ta Lee,§ and Younan Xia*,†,‡,§ †

School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, Georgia 30332, United States § School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia 30322, United States ‡

S Supporting Information *

ABSTRACT: Palladium nanocubes enriched with atomic steps and terraces on the side faces were synthesized at room temperature in an aqueous solution containing a salt precursor, ascorbic acid, KBr, and poly(vinylpyrrolidone) under UV−vis irradiation. The key to the success of this synthesis was the use of PdBr42− as a precursor with strong absorption in the region of 300−600 nm. Single-crystal seeds were formed in the reaction system upon exposure to a medium pressure Hg arc lamp, and the seeds then grew into cubic nanocrystals under extended irradiation. Because the adatoms inherently diffuse slowly across the surface at room temperature, the nanocrystals favored an asymmetric growth mode, generating side faces covered by a series of atomic steps and terraces rather than smooth {100} facets. The steps and terraces would disappear to generate smooth side faces if the sample was subjected to postsynthesis annealing at 130 °C. When applied as catalysts toward the electrochemical oxidation of formic acid, the low-coordination surface atoms such as those presented on the atomic steps resulted in significant reduction of activity relative to the smooth {100} facets. This trend could be rationalized through the overbinding of reactive species to the low-coordination sites.

■

INTRODUCTION Palladium is an essential catalytic material for a variety of industrially important reactions ranging from hydrogenation and dehydrogenation to organic cross-coupling.1−4 Most of these applications benefit from the availability of Pd nanocrystals with well-controlled facets because of their inherently high surface area to volume ratios and the presence of surface structures engineered to suit different chemical reactions. Over the past two decades, an array of methods has been developed for controlling the facets expressed on noble-metal nanocrystals, allowing for the design of heterogeneous catalysts with enhanced activity or selectivity. The synthetic methods can be broadly divided into two major categories: (i) one-pot synthesis in which an organometallic or salt precursor is decomposed or reduced to generate metal atoms, followed by their nucleation and growth into nanocrystals5−8 and (ii) seed-mediated growth, a two-step process wherein seeds are formed in one synthesis and then used in a second process to obtain nanocrystals with a specific shape.9−12 Both approaches benefit from the addition of a capping agent to direct the development of a specific type of facet on the surface. Seed-mediated growth is particularly versatile and viable in generating nanocrystals with complex shapes enclosed by high-index facets. However, multiple steps are often necessary to achieve a tight control over the faceting process. © 2017 American Chemical Society

Most of the synthetic methods reported in literature require the use of an elevated temperature to decompose a precursor or activate its reduction. Thermally-activated processes may experience complications arising from the inhomogeneous distribution of reagents (e.g., the reducing agent or precursor) upon their introduction, as well as the intrinsic temperature gradient in the reaction medium caused by the release, intake, and transfer of heat by molecular species involved in the synthesis. It would thus be highly desirable to develop a roomtemperature method to eliminate these concerns while still maintaining a tight control over the faceting process. In browsing the literature, we only found a small number of reports on the one-step synthesis of Pd nanocrystals at room temperature, the majority of which demonstrated little or no control over the shape (or size) of the nanocrystals.13−15 Part of the problem can be attributed to the limited choice of reduction reactions that can be conducted at room temperature with tunable and controllable kinetics to meet the requirements for both nucleation and growth steps. One particularly attractive solution to this problem is built around photochemical reduction. Compared with a thermally activated Received: March 29, 2017 Revised: April 26, 2017 Published: April 26, 2017 4563

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials

commercial sources, and is substantially more cost-effective to use than its bromo-substituted counterpart, K2PdBr4. The solution was added into a glass vial that already contained 8 mL of an aqueous solution of 105 mg of PVP, 60 mg of AA, 185 mg of KCl, and 400 mg of KBr. The vial was then sealed and placed against a double-walled quartz immersion well in a water bath. The light source (medium pressure Hg vapor arc lamp, Ace Glass) was hosted inside the quartz immersion well. The distance between the light source and the reaction solution was approximately 2 cm. Temperatures of both the water bath and the circulating water from the immersing well were maintained at 23 ± 3 °C throughout the synthesis by controlling the flow rate of water through the immersion well. The reaction was then magnetically stirred under UV−vis irradiation for 3 h. The final product was collected via centrifugation at 55 000 rpm for 30 min, washed three times with water to remove excess PVP, and then redispersed in 11 mL of water. A control reaction was also run under the same conditions except for the use of pure K2PdBr4 instead of Na2PdCl4 as the precursor, and the same experimental results were confirmed. Thermal Transformation of Pd Nanocubes. For a typical thermal transformation test conducted at 80 °C, the Pd nanocubes were heated under magnetic stirring at 80 °C for 3 h. The suspension was then cooled to room temperature and centrifuged at 55 000 rpm for 30 min. The collected pellet was redispersed in 11 mL of water for further analysis. For thermal transformation tests conducted at 130 °C, the Pd nanocubes were first transferred into 11 mL of EG. For experiments involving excess KBr, 400 mg of KBr was also added to the EG suspension and allowed to dissolve through sonication. The samples were then heated under magnetic stirring at 130 °C for 3 h. After cooling to room temperature, the final product was collected by centrifuging the sample at 55 000 rpm for 30 min and washed twice with water. The pellet was collected and redispersed in 11 mL of water for further analysis. Formic Acid Oxidation (FAO). To prepare nanocrystal samples for FAO measurements, samples were attached to a carbon support and the surfaces were cleaned of remnant PVP and Br− through an acetic acid treatment. Ketjen black carbon support (EC-300J) was first suspended in water to a concentration of 1 mg mL−1 by sonicating in ice water for 3 h. To create the ink for electrocatalysis, the selected Pd nanocrystal sample was dropwise added into a carbon suspension to achieve 40% Pd loading by weight. The ink was sonicated for 2 h on ice water. The suspension was then centrifuged at 12 000 rpm for 20 min to obtain a pellet. To clean the surface of the Pd nanocrystals attached to the carbon support, we suspended the pellet into acetic acid and heated the sample under magnetic stirring for 3 h at 60 °C. After the acetic acid treatment, the ink was centrifuged at 12 000 rpm for 20 min to collect a pellet and washed seven times with ethanol and three times with water. The pellet was dispersed in a mixture containing DI water, isopropanol (IPA), and Nafion (5%) (Vwater/ VIPA/V5%Nafion = 4:1:0.05) to form a 1 mg mL−1 ink under ultrasonication for 30 min. The concentration of Pd in the final ink dispersion was determined by inductively coupled plasma mass spectrometry (ICP-MS). Ten microliters of the suspension was then dropped onto a precleaned glassy carbon electrode (Pine Research Instrumentation) and dried at room temperature and under ambient conditions. A Hydroflex hydrogen reference electrode (ET070, Edaq Inc.) and a Pt wire were used as the reference and counter electrodes, respectively. The potentials are given with reference to the reversible hydrogen electrode (RHE). To ensure full removal of any surfactants or capping agents that could adsorbed to the surface of the Pd catalyst (e.g., PVP or Br−), the electrical potential was held at −0.05 V for 60 s. To obtain the FAO activity, the catalyst was cycled between 0.08 and 1.1 V (versus RHE) for two cycles in a N2-saturated aqueous solution that contained HClO4 (0.1 M) and HCOOH (2 M) at a scan rate of 50 mV s−1. The Cu underpotential deposition (UPD) was conducted in an aqueous solution that contained H2SO4 (50 mM) and CuSO4 (50 mM). The electrochemical surface areas (ECSAs) were obtained from the charges associated with the stripping of CuUPD on the Pd nanocrystals by assuming 420 μC cm−2 for a full monolayer coverage of Cu on Pd enclosed by {100} facets. For stability tests of the

process, photochemical reduction offers many potential advantages including homogeneous distribution of reactants throughout the reaction medium, convenient and precise control of energy input, and the ability to maximize the absorption of reactive species by tuning the wavelength of light.16,17 Photochemical reduction can also be conducted at ambient temperature without having to employ an otherwise harsh or hazardous strong reducing agent to generate metal atoms. To our knowledge, most of the research thus far into photochemical synthesis of noble-metal nanocrystals has mainly focused on Au and Ag by leveraging the localized surface plasmon resonance excitation of their nanocrystals in the visible region.17−20 In searching literature, we were only able to identify a few reports on the use of photochemical reduction for the synthesis of Pd nanomaterials.21−25 Most of the papers,21−24 however, only reported photochemical deposition of Pd films or small particles on oxide substrates such as silica, quartz, or titania. The reactions were not maintained at room temperature for these syntheses, and in the limited cases where electron microscopy characterization was provided to support the claims, the Pd nanoparticles appeared to be less than 5 nm in size, with no apparent control over the shape or faceting. Only one of the papers reported the synthesis of colloidal Pd nanocubes and nanorods by irradiating an aqueous mixture of (NH4)2PdCl4, cetyltrimethylammonium bromide (CTAB), and ascorbic acid with 260 nm light for 4 h.25 However, the authors did not mention if the reaction system was maintained at room temperature during UV irradiation, nor did they investigate the mechanism(s) responsible for reduction and particle formation. As shown by the high-resolution transmission electron microcopy (HRTEM) images, the Pd nanocubes had a cubic shape and were covered by smooth side faces, suggesting the possible elevation of reaction temperature during the UV irradiation. Here we report a photochemical route that can be conducted in one pot, at room temperature, and in an aqueous solution for the facile synthesis of Pd nanocubes enriched with atomic steps and terraces on the side faces. Our strategy for controlling the synthesis is multifold, including the use of three major components: (i) Br− as both a selective capping agent toward the {100} facets and as a coordination ligand toward Pd(II) ions; (ii) UV−vis light to activate the Pd(II) precursor; and (iii) ascorbic acid as a chemical reductant. Under optimized conditions, we are able to overcome the energy barrier to initiate reduction at room temperature and controllably generate Pd nanocubes decorated with atomic steps and terraces that cannot otherwise be obtained using a synthesis conducted at an elevated temperature.

■

EXPERIMENTAL SECTION

Chemicals and Materials. Sodium tetrachloropalladate(II) (Na2PdCl4, 98%), potassium bromide (KBr, > 99%), potassium chloride (KCl, > 99%), L-ascorbic acid (AA, > 99%), poly(vinylpyrrolidone) (PVP, Mw ≈ 55 000), and ethylene glycol (EG) were all purchased from Sigma-Aldrich and used as received. Ethanol (200 proof ACS grade) was obtained from VWR. All aqueous solutions were prepared using deionized water with a resistivity of 18.2 MΩ cm. All syntheses were carried out in borosilicate vials (6 dram, VWR, transparent >300 nm) under vigorous magnetic stirring. Photochemical Synthesis of Pd Nanocubes. For a typical synthesis, 57 mg of Na2PdCl4 was dissolved into 3 mL of water over 2 h of vigorous shaking or stirring. We chose to use Na2PdCl4 as our metal salt precursor for this reaction, as it can be readily obtained from 4564

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials

Figure 1. Schematic illustration showing the formation of Pd nanocubes enriched with atomic steps and terraces at room temperature through a photochemical process. The PdBr42− precursor is energized to an excited state by the UV−vis light and then reduced to Pd atoms by ascorbic acid (AA). The Pd atoms nucleate to generate single-crystal seeds, followed by their evolution into truncated cubes and then cubic nanocrystals with atomic steps and terraces on the side faces. photochemically synthesized Pd catalyst, the catalyst electrode was held at 0.45 V in an N2-saturated solution containing HClO4 (0.1 M) and HCOOH (2 M) for 1 h. For stability tests in an oxygen-rich environment, the catalyst electrode was immersed into an O2-saturated solution containing HClO4 (0.1 M) and HCOOH (2 M) for 1 h. All the electrochemical measurements were conducted with a CHI600E potentiostat (CH Instrument). Instrumentation. Samples for TEM were prepared by placing a drop of the nanocrystal sample (suspended in water) on a carboncoated copper grid and drying under ambient conditions. TEM images were taken using a Hitachi HT 7700 microscope operated at 120 kV. Samples for HRTEM were prepared by further washing the sample with water another 5−7 times and then drop casting the sample onto a holey carbon-coated copper grid with ultrathin carbon. HRTEM imaging was performed using a cold field-emission Hitachi HD 2700 aberration-corrected microscope operated at 200 kV. Atomic resolution image for the Supporting Information was acquired using cold field-emission JEOL 4000EX aberration-corrected microscope operated at 350 kV. The UV−vis spectra were obtained using a spectrometer from PerkinElmer (Lambda 750). ICP-MS (NexION 300Q, PerkinElmer) was used to determine the Pd content in each sample.

Figure 2. UV−vis spectra recorded from aqueous solutions of all the reactants involved in the photochemical synthesis. Notably, only the Pd(II) species absorb photons in the region above 300 nm. The spectra were recorded using quartz cuvettes, whereas the photochemical syntheses were conducted in borosilicate glass vials with an absorption cutoff around 300 nm.

■

RESULTS AND DISCUSSION Figure 1 shows a general schematic of our synthesis, which involves the use of both AA and UV−vis light to reduce Pd(II) species to Pd(0) atoms at room temperature. The Pd(0) atoms then nucleate to generate single-crystal seeds, followed by their continuous growth into nanocrystals enclosed by {100} facets owing to the presence of Br− as a selective capping agent toward Pd(100). Many different pathways have been reported in the literature to account for the photochemical reduction of transition metal ions, including but not limited to direct photochemical reduction through irradiation, as well as reduction through the use of a molecular sensitizer. In particular, many photochemical methods rely on the use of an organic dye or chromophore that will decompose to generate radicals upon irradiation at an appropriate wavelength. These highly reactive radical species can readily reduce transition metal ions into atoms, driving the formation of nanoparticles.17,26,27 Initially, we believed this reduction mechanism might be involved in our synthesis as AA could potentially be cleaved to form an ascorbyl radical.28 However, the synthesis was conducted in water and in the presence of O2 from air. Molecular oxygen is known to be an excellent radical scavenger that can easily quench radicals, thus making a radicalbased reduction pathway unlikely. To verify this argument, we measured the UV−vis absorption spectra of all the individual components involved in our synthesis (Figure 2). While AA does possess a sharp absorption peak below 300 nm, the borosilicate glass vial used to house the reaction solution should

be able to block all irradiation below 300 nm. Therefore, we can discard a reduction mechanism based on the ascorbyl radical. Instead, the reduction pathway critical to the success of this synthesis relies on the coupling between AA and UV−vis irradiation, which is highly dependent on what Pd(II) species is actually present in the reaction solution. Bromide is well-known for selectively capping Pd{100} facets and has been extensively used to promote the formation of {100}-dominated nanocrystals such as cubes by effectively blocking the deposition of Pd new atoms on {100} facets, and thereby promoting ⟨111⟩ growth instead. However, the role of Br− in our synthesis is two-fold, starting before Pd nanocrystals are even formed. In an aqueous solution containing halide salts, Pd(II) can form a square-planar complex coordinated by halides (Cl− or Br−, in our reaction solution) or by water.29−31 The respective d-band and ligand−metal charge transfer absorption peaks for PdBrn(H2O)4−n2−n and PdCln(H2O)4−n2−n (where n ≤ 4) have been well established in the literature and are distinct for each hydrated species.32,33 Because of the difference in the respective stability constants, the exact ligands present in the coordination complex will have a strong impact on the reduction kinetics.32,34,35 A more highly Br−-substituted complex will yield a slower reduction rate because of the increase in compound stability. For example, PdCl42− can be readily reduced in the presence of AA within a few minutes, 4565

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials even at room temperature. In contrast, our reaction mixture could not be reduced at all while sitting at room temperature in the absence of UV−vis irradiation. To this end, it was critical to identify the predominant Pd(II) species present in the reaction solution. Figure S1 shows a UV−vis absorption spectrum of the reaction solution in the absence of AA; all of the peaks correspond to PdBr42−. Because of the high concentration of excess Br− present in the solution, all of the Pd(II) species became PdBr42− through a quick ligand exchange process. It was PdBr42− that served as a salt precursor for the actual reduction and subsequent nanocrystal nucleation and growth, even when the initial precursor added into the reaction system was Na2PdCl4. If there was not sufficient excess of Br− present in the reaction solution to fully substitute the Cl− ligand, it would change the stability constant of Pd(II) complex and the reduction kinetics, potentially resulting in uncontrolled nucleation and loss of control over facet expression. To achieve control over both nucleation and growth, it is critical not only to develop a reduction process that is fast enough in the initial stage to ensure the formation of singlecrystal seeds, but also sufficiently slow to allow for facetcontrolled growth. As stated previously, PdBr42− is a stable compound that cannot be readily reduced by AA at room temperature. While there are other stronger reducing agents that can act at room temperature (i.e., NaBH4 and hydrazine), they can reduce PdBr42− too quickly and eliminate any facet control in the growth stage. Additionally, many of these reducing agents are hazardous with which to work. We addressed this issue by employing a combination of AA and UV−vis light. Upon irradiation by a medium-pressure Hg arc lamp with strong outputs between 300−600 nm (see Figure S2), the PdBr42− complexes can absorb an adequate number of photons to reach an excited state and generate [PdBr42−]*. At this excited state, the Pd(II) precursor can be readily reduced by AA even at room temperature to generate Pd(0) atoms. To ensure that the reduction is not caused by the higher energy photons alone, we confirmed that the reaction did not occur in the presence of UV−vis irradiation unless AA was also present. This observation verifies that the synthesis was based on the direct reduction of [PdBr42−]* by AA. Only under these conditions could controlled nucleation and facet-directed growth occur at room temperature. Once the Pd(0) atoms had reached a high enough concentration, they would undergo a classical nucleation process. Atomic resolution TEM from a typical reaction after 10 min of UV−vis irradiation revealed the presence of small, single-crystal seeds (Figure S3a). After 45 min of irradiation, the nucleates began to undergo facet-directed growth, generating small cuboctahedral crystals due to the presence of Br− to cap the {100} faces (Figure S3b). Upon 2 h of irradiation, the cuboctahedra further grew into larger cubes, more closely resembling the nanocrystals in a product collected after a full, 3-h synthesis (Figure S3c). However, in contrast to the relatively sharp cubes produced through thermal routes,36 the Pd nanocrystals formed in the present method, while cubelike, were also notably irregular and nonuniform, demonstrating the dominance of an asymmetric growth mode (Figure 3a), even though the XRD pattern confirmed the formation of pure Pd crystals (Figure 3B). HRTEM was used to resolve the exact morphology of the final Pd nanocrystals produced through the photochemical method. Figure 4A shows a high-resolution image of a typical Pd nanocrystal. While the corners are truncated and rounded

Figure 3. (A) TEM image of the Pd nanocrystals obtained from a typical photochemical synthesis, indicating a mixture of regular, symmetric cubes “S” with smooth side faces and asymmetric, cube-like particles “A” enriched with atomic steps and terraces. The distributions in the inset were calculated by examining 300 particles in a typical sample. Since the statistics were based on TEM images, the 18% should be considered as an upper limit for the symmetric cubes. (B) XRD pattern of a typical nanocrystal sample, matching a pure fcc Pd spectrum (JCPDS card No. 05−0681).

rather than being sharp, the side faces of the nanocrystals are of particular interest. Atomic resolution images (Figure 4B,C) of the areas enclosed by red boxes in Figure 4A reveal that the side faces are not composed of the smooth {100} facets expected for a perfect nanocube, but instead form an irregular, almost concave structure that cannot be readily assigned to a single, high-index facet. Rather, the faces along the edges of the nanocrystal are composed of a series of random atomic steps and terraces at a decline normal to the (100) face. The observed asymmetric growth pattern can be better understood by examining the reaction conditions of the photochemical synthesis as compared to the traditional thermal method and the difference in the relative rates for atom deposition (Vdep) and adatom surface diffusion (Vdif).37,38 It has been demonstrated that the rate of atom deposition can be 4566

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials

greatly reduced compared to the case at an elevated temperature used for the thermally activated synthesis. In a recent study, we demonstrated that this drastically reduced Vdif at room temperature directly inhibits the ability of the newly deposited Pd adatoms to diffuse across the surface of a seed, limiting the active sites for growth to some of the corners and further inducing an asymmetric growth mode.42 As a result, the photochemical process always favors asymmetric growth, resulting in the formation of asymmetric and truncated cubes as shown in a typical synthesis. Furthermore, the sluggish Vdif slows diffusion of new adatoms from the {111} corners where they are deposited to the Br− capped {100} faces. This in turn results in the high percentage of atomic steps and terraces observed under atomic resolution TEM (Figure 4), rather than smooth {100} facets. To generate more uniform cubes with a symmetric growth mode, the photochemical reduction process would need to be conducted at an elevated temperature, promoting faster adatom surface diffusion relative to the deposition process. We also investigated the effect of postsynthetic heat treatment on the asymmetric Pd cubes in an attempt to determine if the shape could be controllably altered under thermal annealing. It was found that the highly terraced faces on the original Pd particles could be transformed during thermal treatment under appropriate conditions (Figure 6). In an effort to understand the stability of these irregular facets, we heated the Pd nanocrystals with atomic steps and terraces to 80 °C in water, the typical temperature used to synthesize sharp Pd cubes. However, even after 3 h of heating, the original nanocrystals remained stable and maintained the same

Figure 4. (A) HRTEM image of a typical Pd nanocrystal produced using the photochemical route. (B, C) Atomic-resolution TEM images of the faces marked by red boxes in panel A, demonstrating the enrichment of atomic steps and terraces on the side faces.

controlled by tuning the reduction rate of metal ions in a reaction solution.39−41 In contrast, Vdif is strongly dependent on the temperature of a synthesis, as illustrated in Figure 5. At a temperature of 80 °C typically used for the thermal synthesis of Pd nanocubes, Vdif is adequate to transport the newly deposited Pd adatoms across the {100} facets as the cuboctahedral seeds continue to grow in size. At room temperature, however, Vdif is

Figure 6. TEM images showing the effect of postsynthesis thermal annealing on the shape of the Pd cubes with atomic steps and terraces. (A) Initial nanocrystals synthesized through the standard photochemical protocol. (B) Nanocrystals from panel A after annealing in water at 80 °C for 3 h, showing preservation of shape and morphology. (C) Nanocrystals from panel A after annealing in ethylene glycol at 130 °C for 3 h, showing a loss of the initial cubic shape and atomic steps and terraces as surface diffusion of adatoms readily occurs, leading to the formation of multifaceted Wulff structures with a lower total surface free energy. (D) Nanocrystals from panel A after annealing in ethylene glycol with excess KBr at 130 °C for 3 h, showing their transformation to truncated cubes with smooth faces.

Figure 5. Schematic illustration demonstrating the effect of reaction temperature on the exact morphology taken by the Pd nanocrystals. At elevated temperatures, the rate of surface diffusion (Vdif) is high enough to move adatoms from the deposition sites across the {100} facets, producing symmetric cubes with sharp corners and smooth side faces. At room temperature, Vdif is substantially reduced, such that adatoms cannot readily diffuse across the side faces of a cube, resulting in the formation of Pd nanocubes enriched with atomic steps and terraces. 4567

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials

Figure 7. TEM images of different types of Pd nanocrystals that were used as catalysts for formic acid oxidation, and their catalytic results. (A) Pd nanocubes with sharp corners and smooth faces that were prepared using the traditional thermal method. (B) Pd nanocubes with atomic steps and terraces that were synthesized through the photochemical route. (C) Pd nanocubes with truncated corners and smooth faces that were obtained via thermal transformation of panel B at 130 °C in the presence of excess KBr. The scale bar is the same for panels A−C. (D) CVs of the catalysts obtained in a N2-saturated aqueous solution containing H2SO4 and CuSO4. Scanning speed = 20 mV s−1. The currents were normalized to the geometric area of the glassy carbon electrode (0.196 cm2). (E) CVs of the catalysts obtained in a N2-saturated solution containing 0.1 M HClO4 and 0.5 M HCOOH. Scanning speed = 50 mV s−1. The currents were normalized to the electrochemical surface areas of the corresponding catalysts derived from the charges for the underpotential deposition of Cu.

Table 1. Comparison of Formic Acid Oxidation Activities for Catalysts Based on Different Types of Pd Nanocrystals Shown in Figure 7A−C catalyst

specific ECSA (m2 g−1Pd)

SA at 0.4 V (mA cm−2)

MA at 0.4 V (A mg−1Pd)

anodic peak potential (V)

SA at anodic potential (mA cm−2)

MA at anodic potential (A mg−1Pd)

1 2 3

11.38 ± 1.10 13.38 ± 2.39 7.43 ± 1.02

1.93 ± 0.15 0.67 ± 0.10 1.10 ± 0.17

0.22 ± 0.01 0.09 ± 0.002 0.08 ± 0.001

0.52 ± 0.01 0.46 ± 0.01 0.50 ± 0.01

2.90 ± 0.29 0.79 ± 0.13 1.48 ± 0.26

0.33 ± 0.01 0.11 ± 0.003 0.11 ± 0.001

of low-coordination active sites, the associated energy between the atomic site and the binding species may increase relative to the low index sites, leading to one of two results: the increased binding energy may favorably facilitate reactive species binding and improve the subsequent catalytic reaction, or it may cause the active site to “overbind” to reactant/intermediate species and fall into a thermodynamic sink.48,49 Overbinding results in a dramatic decrease in the measured catalytic performance, similar to catalyst poisoning. Short of employing computationally expensive calculations, determining how high energy active sites will perform requires case-by-case testing. Therefore, to better evaluate the effect of the Pd cubes with atomic steps and terraces on potential applications, we evaluated the catalytic performance using FAO as a model reaction. Prior studies on Pd single crystal catalysts have shown a facet dependence for FAO, with cubes enclosed by {100} facets displaying enhanced activity over their octahedral counterparts enclosed by {111} facets.50 For a standard comparison, we synthesized 12 nm Pd cubes (Figure 7A) with sharp corners and smooth faces through a traditional thermal method. We then compared the sharp 12 nm cubes to a typical sample of Pd cubes with atomic steps and terraces (Figure 7B) and to a sample of truncated cubes with smooth faces (Figure 7C) formed through the thermal transformation of the photochemical cubes.

morphology, showing no sign of surface atom diffusion (Figure 6B). The nanocrystals were transferred into ethylene glycol and heated at 130 °C for 3 h. At this temperature, the nanocrystals underwent surface adatom diffusion and structural transformation; however, they evolved into spherical, Wulff-like structures with a mixture of various low-index facets (Figure 6C). The Wulff structure is a morphology that most singlecrystal metals are known to adopt to minimize the total surface free energy when no capping agents are present to direct and stabilize growth. In an effort to generate sharp cubes, we conducted the same heating process at 130 °C for 3 h but in the presence of excess Br−. With excess Br− available to selectively cap the {100} facets and stabilize the cubic shape, we were able to induce surface transformation into truncated cubes (Figure 6D). The Pd adatoms initially present on the atomic steps and terraces were able to readily diffuse across the surface, forming cubes with rounded corners but smooth faces. Many studies have suggested that the high surface energy associated with low-coordination atomic sites may contribute toward improved catalytic performance.43−45 However, this structure−property relationship is not well-defined and is highly dependent on both the catalytic material and the targeted reaction. Computational modeling is often required to understand the energetics of the binding between the active sites and the catalytic reactants or intermediates.46,47 In the case 4568

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials

tions for future work done on nanocatalysts with a high proportion of low-coordination active sites.

The results of the electrochemical measurements are shown in Figure 7D and E. Figure 7D shows the cyclic voltammograms (CVs) of the three catalytic samples for Cu underpotential deposition. The ECSAs of the catalysts were determined from the charges associated with stripping a UPD monolayer of Cu from the catalyst surfaces. The specific ECSAs were then calculated by normalizing the measured ECSAs to the mass of Pd in each sample; all calculated values are summarized in Table 1. Figure 7E shows the cyclic voltammograms of the catalysts for FAO, normalized to their respective ECSAs. As the anodic peak potential for FAO can shift depending on the surface structure of the nanocrystals involved, we calculated and compared the specific activities (SAs) at the respective anodic peak potentials for each sample. The FAO measurements yielded a clear trend in the calculated SAs. The 12 nm Pd sharp cubes yielded the highest SA (2.90 mA cm−2 at anodic peak potential). In contrast, the Pd cubes with atomic steps and terraces demonstrated the poorest SA (0.79 mA cm−2 at anodic peak potential, 3.6 times lower than the sharp cubes). Finally, the thermally transformed photochemical cubes with smooth faces, but truncated corners yielded a calculated SA that was higher than the terraced cubes but lower than the cubes with sharp corners (1.48 mA cm−2 at anodic peak potential, 1.9times higher than the terraced cubes and 1.9-times lower than the sharp cubes). This trend can be rationalized by the differences in the surface and electronic structures between the three catalytic samples. Contrary to literature that has reported enhanced FAO activity on some nanocrystals with high index facets,51,52 here we believe the high proportion of lowcoordinated surface atoms from the atomic steps and terraces is directly responsible for the large drop in the measured peak anodic SA as compared to the sharp cubes. Interestingly, when the photochemically synthesized catalysts are placed into oxidative environments (oxygen saturated electrolyte, or held at FAO oxidative potentials), there appear to be no dramatic or immediate changes to the irregular surface structure (Figure S4), even after an hour of oxidative stress. This suggests that the persistent low coordination sites present on the faces are largely responsible for the significant decrease in observed FAO activity. Rather than enhancing the performance, the altered electronic structure of these high energy sites instead appears to decrease the catalytic activity. We believe this is due to overbinding of the active sites with the formic acid or a reactive intermediate, resulting in dramatically reduced catalytic conversion and the apparent drop in catalytic performance. With this in mind, we can therefore attribute the enhancement in SA seen in the thermally transformed cubes to the sharpening of the cubic faces into true {100} facets devoid of atomic steps, ledges, and low-coordination sites. Conversely, though this transformation improves the SA relative to the photochemically synthesized cubes, the true cubes with sharp corners still demonstrate the best catalytic performance. In agreement with earlier studies, this can be rationalized due to the highly truncated corners in the thermally transformed sample.50 While the smooth {100} facets enhance activity, the presence of highly truncated corners and therefore a mixture of both {100} and {111} facets on the nanocatalysts reduces the anodic SA relative to the true, sharp cubes. This lends credit to the experimentally observed pattern wherein single crystal Pd catalysts with the highest proportion of {100} facets exhibit the highest activity toward FAO. These results demonstrate a clear correlation between the surface structure of nanocrystals and their measured catalytic performance, with important implica-

■

CONCLUSION

■

ASSOCIATED CONTENT

In summary, we have successfully demonstrated a facile route to the photochemical synthesis of Pd nanocrystals at room temperature. The key to this synthesis was two-fold by relying on the use of a strongly irradiating light source as well as room temperature reaction conditions. The strong UV−vis irradiation from a Hg lamp was found to be critical for both the generation of seeds as well as their subsequent growth at room temperature. Simultaneously, because of the inherently low surface diffusion rate of Pd adatoms at room temperature, the nanocrystals evolved into truncated cubes with an asymmetrical shape, naturally forming irregular atomic steps and terraces on their faces rather than smooth {100} facets as found at elevated reaction temperatures. This photochemical method opens an avenue toward generating Pd nanocrystals at room temperature and toward investigating novel morphologies characterized by atomic steps and terraces rather than the low-index facets typically generated through high temperature processes. Using formic acid oxidation as a model catalytic reaction, we compared the terraced nanocubes to the traditional nanocubes with smooth {100} facets. Our results indicate a clear trend wherein the catalytic performance was enhanced by the presence of smooth {100} facets, but decreased with an increasing proportion of low-coordination surface atoms. This can be attributed to the overbinding of reactive species to the high energy sites. Contrary to prior reports on the benefits of high-index nanocrystals, our results highlight the importance of fully investigating the impact of low-coordination active sites on catalytic applications.

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.chemmater.7b01287. UV−vis spectra of Pd(II) species; energy distribution spectra of mercury vapor arc lamp; atomic-resolution and conventional TEM images of nanocrystals; HAADFSTEM images of photochemical Pd catalyst loaded onto carbon support (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Madeline Vara: 0000-0001-8968-8734 Younan Xia: 0000-0003-2431-7048 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported in part by a DOE subcontract from the University of Wisconsin−Madison (DE-FG02 05ER15731) and startup funds from the Georgia Institute of Technology. 4569

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials

■

(22) Boitsova, T. B.; Gorbunova, V. V.; Voronin, Y. M. Colloidal palladium: Photochemical production and optical properties. J. Opt. Technol. 2001, 68, 789−793. (23) Lifanova, E. A.; Gorbunova, V. V.; Boitsova, T. B. Photochemical synthesis of nano-phase palladium films on quartz. Russ. J. Appl. Chem. 2009, 82, 915−917. (24) Chen, S. F.; Li, J. P.; Qian, K.; Xu, W. P.; Lu, Y.; Huang, W. X.; Yu, S. H. Large scale photochemical synthesis of M@TiO2 nanocomposites (M = Ag, Pd, Au, Pt) and their optical properties, CO oxidation performance, and antibacterial effect. Nano Res. 2010, 3, 244−255. (25) Kundu, S.; Wang, K.; Lau, S.; Liang, H. Photochemical synthesis of shape-selective palladium nanocubes in aqueous solution. J. Nanopart. Res. 2010, 12, 2799−2811. (26) McGilvray, K. L.; Decan, M. R.; Wang, D.; Scaiano, J. C. Facile Photochemical Synthesis of Unprotected Aqueous Gold Nanoparticles. J. Am. Chem. Soc. 2006, 128, 15980−15981. (27) Scaiano, J. C.; Stamplecoskie, K. G.; Hallett-Tapley, G. L. Photochemical Norrish type I reaction as a tool for metal nanoparticle synthesis: importance of proton coupled electron transfer. Chem. Commun. 2012, 48, 4798−4808. (28) Tikekar, R. V.; Anantheswaran, R. C.; Elias, R. J.; LaBorde, L. F. Ultraviolet-induced Oxidation of Ascorbic Acid in a Model Juice System: Identification of Degradation Products. J. Agric. Food Chem. 2011, 59, 8244−8248. (29) Elding, L. I. Kinetics for anation of aqua palladate(II) complexes. Inorg. Chim. Acta 1975, 15, L9−L11. (30) Richens, D. T. Ligand Substitution Reactions at Inorganic Centers. Chem. Rev. 2005, 105, 1961−2002. (31) Timoshkin, A. Y.; Kudrev, A. G. Calculations of microconstants and equilibrium formation constants for platinum(II) and palladium(II) halide complexes in solution. Russ. J. Inorg. Chem. 2012, 57, 1362− 1370. (32) Elding, L. I. Palladium(II) halide complexes. I. Stabilities and spectra of palladium(II) chloro and bromo aqua complexes. Inorg. Chim. Acta 1972, 6, 647−651. (33) Elding, L. I.; Olsson, L. F. Electronic absorption spectra of square-planar chloro-aqua and bromo-aqua complexes of palladium(II) and platinum(II). J. Phys. Chem. 1978, 82, 69−74. (34) Srivastava, S. C.; Newman, L. Mixed Ligand Complexes of Palladium(II) with Chloride and Bromide. Inorg. Chem. 1966, 5, 1506−1510. (35) Feldberg, S.; Klotz, P.; Newman, L. Computer evaluation of equilibrium constants from spectrophotometric data. Inorg. Chem. 1972, 11, 2860−2865. (36) Lim, B.; Jiang, M.; Tao, J.; Camargo, P. H. C.; Zhu, Y.; Xia, Y. Shape-controlled Synthesis of Pd Nanocrystals in Aqueous Solutions. Adv. Funct. Mater. 2009, 19, 189−200. (37) Xia, Y.; Xia, X.; Peng, H.-C. Shape-controlled Synthesis of Colloidal Metal Nanocrystals: Thermodynamic versus Kinetic Products. J. Am. Chem. Soc. 2015, 137, 7947−7966. (38) Zhu, C.; Zeng, J.; Tao, J.; Johnson, M. C.; Schmidt-Krey, I.; Blubaugh, L.; Zhu, Y.; Gu, Z.; Xia, Y. Kinetically Controlled Overgrowth of Ag or Au on Pd Nanocrystal Seeds: From Hybrid Dimers to Nonconcentric and Concentric Bimetallic Nanocrystals. J. Am. Chem. Soc. 2012, 134, 15822−15831. (39) Biacchi, A. J.; Schaak, R. E. The Solvent Matters: Kinetic versus Thermodynamic Shape Control in the Polyol Synthesis of Rhodium Nanoparticles. ACS Nano 2011, 5, 8089−8099. (40) Xie, S.; Choi, S.-I.; Lu, N.; Roling, L. T.; Herron, J. A.; Zhang, L.; Park, J.; Wang, J.; Kim, M. J.; Xie, Z.; Mavrikakis, M.; Xia, Y. Atomic Layer-by-Layer Deposition of Pt on Pd Nanocubes for Catalysts with Enhanced Activity and Durability toward Oxygen Reduction. Nano Lett. 2014, 14, 3570−3576. (41) Zhang, L.; Niu, W.; Xu, G. Synthesis and applications of noble metal nanocrystals with high-energy facets. Nano Today 2012, 7, 586− 605.

REFERENCES

(1) Shu, J.; Grandjean, B. P. A.; Neste, A. V.; Kaliaguine, S. Catalytic palladium-based membrane reactors: A review. Can. J. Chem. Eng. 1991, 69, 1036−1060. (2) Miyaura, N.; Suzuki, A. Palladium-Catalyzed Cross-Coupling Reactions of Organoboron Compounds. Chem. Rev. 1995, 95, 2457− 2483. (3) Kwon, M. S.; Kim, N.; Park, C. M.; Lee, J. S.; Kang, K. Y.; Park, J. Palladium Nanoparticles Entrapped in Aluminum Hydroxide: Dual Catalyst for Alkene Hydrogenation and Aerobic Alcohol Oxidation. Org. Lett. 2005, 7, 1077−1079. (4) Yun, S.; Ted Oyama, S. Correlations in palladium membranes for hydrogen separation: A review. J. Membr. Sci. 2011, 375, 28−45. (5) Kim, S.-W.; Park, J.; Jang, Y.; Chung, Y.; Hwang, S.; Hyeon, T.; Kim, Y. W. Synthesis of Monodisperse Palladium Nanoparticles. Nano Lett. 2003, 3, 1289−1291. (6) Ramirez, E.; Jansat, S.; Philippot, K.; Lecante, P.; Gomez, M.; Masdeu-Bultó, A. M.; Chaudret, B. Influence of organic ligands on the stabilization of palladium nanoparticles. J. Organomet. Chem. 2004, 689, 4601−4610. (7) Piao, Y.; Jang, Y.; Shokouhimehr, M.; Lee, I. S.; Hyeon, T. Facile Aqueous-Phase Synthesis of Uniform Palladium Nanoparticles of Various Shapes and Sizes. Small 2007, 3, 255−260. (8) Xiong, Y.; Xia, Y. Shape-Controlled Synthesis of Metal Nanostructures: The Case of Palladium. Adv. Mater. 2007, 19, 3385−3391. (9) Niu, W.; Zhang, L.; Xu, G. Shape-Controlled Synthesis of SingleCrystalline Palladium Nanocrystals. ACS Nano 2010, 4, 1987−1996. (10) Wang, Y.; Xie, S.; Liu, J.; Park, J.; Huang, C. Z.; Xia, Y. Shapecontrolled Synthesis of Palladium Nanocrystals: A Mechanistic Understanding of the Evolution from Octahedrons to Tetrahedrons. Nano Lett. 2013, 13, 2276−2281. (11) Peng, H.-C.; Li, Z.; Aldahondo, G.; Huang, H.; Xia, Y. SeedMediated Synthesis of Pd Nanocrystals: The Effect of Surface Capping on the Heterogeneous Nucleation and Growth. J. Phys. Chem. C 2016, 120, 11754−11761. (12) Xia, Y.; Gilroy, K. D.; Peng, H.-C.; Xia, X. Seed-Mediated Growth of Colloidal Metal Nanocrystals. Angew. Chem., Int. Ed. 2017, 56, 60−95. (13) Mu, X.-d.; Evans, D. G.; Kou, Y. A General Method for Preparation of PVP-Stabilized Noble Metal Nanoparticles in Room Temperature Ionic Liquids. Catal. Lett. 2004, 97, 151−154. (14) Nadagouda, M. N.; Varma, R. S. Green synthesis of silver and palladium nanoparticles at room temperature using coffee and tea extract. Green Chem. 2008, 10, 859−862. (15) Fan, F.-R.; Attia, A.; Sur, U. K.; Chen, J.-B.; Xie, Z.-X.; Li, J.-F.; Ren, B.; Tian, Z.-Q. An Effective Strategy for Room-Temperature Synthesis of Single-Crystalline Palladium Nanocubes and Nanodendrites in Aqueous Solution. Cryst. Growth Des. 2009, 9, 2335− 2340. (16) Grzelczak, M.; Liz-Marzan, L. M. The relevance of light in the formation of colloidal metal nanoparticles. Chem. Soc. Rev. 2014, 43, 2089−2097. (17) Sakamoto, M.; Fujistuka, M.; Majima, T. Light as a construction tool of metal nanoparticles: Synthesis and mechanism. J. Photochem. Photobiol., C 2009, 10, 33−56. (18) Kim, F.; Song, J. H.; Yang, P. Photochemical Synthesis of Gold Nanorods. J. Am. Chem. Soc. 2002, 124, 14316−14317. (19) Callegari, A.; Tonti, D.; Chergui, M. Photochemically Grown Silver Nanoparticles with Wavelength-Controlled Size and Shape. Nano Lett. 2003, 3, 1565−1568. (20) Millstone, J. E.; Hurst, S. J.; Métraux, G. S.; Cutler, J. I.; Mirkin, C. A. Colloidal Gold and Silver Triangular Nanoprisms. Small 2009, 5, 646−664. (21) Borgarello, E.; Serpone, N.; Emo, G.; Harris, R.; Pelizzetti, E.; Minero, C. Light-induced reduction of rhodium(III) and palladium(II) on titanium dioxide dispersions and the selective photochemical separation and recovery of gold(III), platinum(IV), and rhodium(III) in chloride media. Inorg. Chem. 1986, 25, 4499−4503. 4570

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571

Article

Chemistry of Materials (42) Peng, H.-C.; Park, J.; Zhang, L.; Xia, Y. Toward a Quantitative Understanding of Symmetry Reduction Involved in the Seed-Mediated Growth of Pd Nanocrystals. J. Am. Chem. Soc. 2015, 137, 6643−6652. (43) Tian, N.; Zhou, Z.-Y.; Sun, S.-G.; Ding, Y.; Wang, Z. L. Synthesis of Tetrahexahedral Platinum Nanocrystals with High-Index Facets and High Electro-Oxidation Activity. Science 2007, 316, 732− 735. (44) Tian, N.; Zhou, Z.-Y.; Yu, N.-F.; Wang, L.-Y.; Sun, S.-G. Direct Electrodeposition of Tetrahexahedral Pd Nanocrystals with HighIndex Facets and High Catalytic Activity for Ethanol Electrooxidation. J. Am. Chem. Soc. 2010, 132, 7580−7581. (45) Xie, X.; Gao, G.; Pan, Z.; Wang, T.; Meng, X.; Cai, L. Largescale Synthesis of Palladium Concave Nanocubes with High-Index Facets for Sustainable Enhanced Catalytic Performance. Sci. Rep. 2015, 5, 8515. (46) Falicov, L. M.; Somorjai, G. A. Correlation between catalytic activity and bonding and coordination number of atoms and molecules on transition metal surfaces: Theory and experimental evidence. Proc. Natl. Acad. Sci. U. S. A. 1985, 82, 2207−2211. (47) Greeley, J.; Nørskov, J. K.; Mavrikakis, M. Electronic Structure and Catalysis on Metal Surfaces. Annu. Rev. Phys. Chem. 2002, 53, 319−348. (48) Liu, C.; Cundari, T. R.; Wilson, A. K. CO2 Reduction on Transition Metal (Fe, Co, Ni, and Cu) Surfaces: In Comparison with Homogeneous Catalysis. J. Phys. Chem. C 2012, 116, 5681−5688. (49) Holby, E. F.; Taylor, C. D. Activity of N-coordinated multimetal-atom active site structures for Pt-free oxygen reduction reaction catalysis: Role of *OH ligands. Sci. Rep. 2015, 5, 9286. (50) Jin, M.; Zhang, H.; Xie, Z.; Xia, Y. Palladium nanocrystals enclosed by {100} and {111} facets in controlled proportions and their catalytic activities for formic acid oxidation. Energy Environ. Sci. 2012, 5, 6352−6357. (51) Hoshi, N.; Nakamura, M.; Kida, K. Structural effects on the oxidation of formic acid on the high index planes of palladium. Electrochem. Commun. 2007, 9, 279−282. (52) Jin, M.; Zhang, H.; Xie, Z.; Xia, Y. Palladium Concave Nanocubes with High-Index Facets and Their Enhanced Catalytic Properties. Angew. Chem., Int. Ed. 2011, 50, 7850−7854.

4571

DOI: 10.1021/acs.chemmater.7b01287 Chem. Mater. 2017, 29, 4563−4571